A method for characterizing an emotional state caused by sensory stimulation.

Abstract

The invention relates to a method for characterizing sensory stimulation, comprising the steps of: subjecting a human subject to visual stimulation, then to the sensory stimulation to be characterized, and to an image showing several emojis; acquiring and processing eye-tracking signals (ETC) during the image presentation to determine the first emoji (EP1) that is fixed for the longest time, this first emoji determining a group of themes likely to be felt during the sensory stimulation; and acquiring and processing physiological measurement signals (FP1S, FP2S, AF7S, AF8S, F3S, F4S, SS1, SS2) during the presentation of the visual and sensory stimulations to identify a theme felt within the group of themes. Figure 5

Description

Title of the invention: Method for characterizing an emotional state induced by sensory stimulation. Technical field

[0001] The present invention relates to the characterization of a theme, in particular an emotional state of a human subject caused by sensory stimulation, such as visual, olfactory, tactile, auditory or gustatory stimulation. State of the art

[0002] Various methods for assessing the perception of a stimulus have already been proposed. For example, self-report questionnaires for identifying emotional states have been proposed (ref. [1]). It is also known to characterize an emotional state using physiological measures, but these do not allow for the discrimination of emotions beyond a positive or negative valence (refs. [2] [Ma et al., 2020]). It has also been proposed to use behavioral measures, such as facial expression analysis, to discriminate facial reactions according to the six basic emotions (Ref. [3]). However, this number of emotions remains very limited, with a majority of negative emotions being dominant.

[0003] Another method uses mood mapping to emotionally classify tested olfactory products (Ref. [4]). This classification is based on participants choosing, from among eight categories, an emotional term that best reflects the emotion induced by the product. According to another method, participants must assign an intensity (from 0 to 10) to each emotional term in a list provided for each product to be evaluated (Ref. [5]). A non-verbal conscious method based on a selection of emotional images has also been proposed, as well as a method using electroencephalography to measure a person's state of relaxation (Ref. [6] described).

[0004] It has also been noted (ref. [7]) that analyzing emotions related to taste perception using questionnaires such as that described in ref. [8] is unconventional and specific, since consumers tend to cognitively associate terms with products without directly experiencing them ([ref. 9]). Emotions are described as short-term affective responses to stimuli with reinforcement potential. Therefore, research is needed to understand how consumers naturally express emotions related to smell or taste.

[0005] Furthermore, the advent of the internet and social media has provided a new medium for the non-verbal expression of emotions through emojis (i.e., (This refers to pictograms symbolizing an emotion). On the X platform (formerly Twitter), users refer to mood and emotions in 25% of tweets, with emojis being used more frequently than words. It turns out that emojis are considered not only a partial substitute for standard language (ref.

[10] ), but also an automatic and effortless way to convey emotions, regardless of gender or age group (refs.

[11] ,

[12] ,

[13] , and

[14] ). In Korea, emojis are the third most used language, after Korean and English (ref. [7]). Thus, emojis have, in a way, revolutionized our modes of communication. When a person uses an emoji, they signify an action resulting from the emotion they felt.

[0006] The use of electroencephalograms (EEGs) is a method commonly employed by researchers to assess the pleasure perceived by a human subject. More specifically, frontal alpha asymmetry (FAA) has been associated with emotion processing and affective psychopathology. High levels of right cortical activity have been associated with negative and withdrawal stimuli and behaviors, while increased left cortical activity has been associated with positive and approach stimuli and behaviors (ref.

[15] ).

[0007] Eye-tracking technology is frequently used to assess cognitive and decision-making processes (ref.

[16] ,

[17] ,

[18] ). Eye tracking can be correlated with the selection of emojis to express feelings, thus revealing the decision-making process. Eye tracking is also commonly used in various emotion recognition methods (ref.

[19] ).

[0008] To analyze emotions, it has also been proposed to integrate three components, namely: the physiological component, which defines the intensity of emotions; the subjective component, which determines the nature of emotions; and the behavioral component, which identifies actions in response to emotions. However, this method is only effective if these three components can be integrated without complicating the experimental protocol (ref.

[20] .

[0009] All the methods mentioned above for characterizing an emotional state do not allow for a full analysis of the emotional responses of human subjects, because the conscious / subjective component predominates. Furthermore, the objective physiological measures used do not allow for the differentiation of felt emotions beyond the valence scale. Behavioral measures, such as facial and vocal analysis, only allow for the study of a limited number of emotions, which may be insufficient to achieve certain objectives.

[0010] To overcome these limitations, the Applicant has proposed a method for characterizing a felt theme elicited by olfactory stimulation, as described in French patent FR3129072 (or US patent application 2023 / 148941). In this In this method, the subject's experience emerges from interaction with a product in a given context. This method examines the perceptions of a group of participants by integrating subjective and physiological components. The subjective component is assessed using questionnaires, while the physiological component involves recording blood volume (BVP) and measuring skin conductance (SCL). Participants' emotions are then calibrated using videos in virtual reality environments. These videos are shown to the participants, who are asked to evaluate them emotionally while physiological measurements are acquired. The emotional term chosen by each participant is then associated with the physiological reactions thus measured.This data is compiled into a database, which is then used to identify the emotions generated in each participant when they smell a product. To this end, the physiological reactions measured during the olfactory perception of a product are compared to those calibrated using virtual reality videos. In this way, the emotion recognized in response to the product is the one whose physiological reactions obtained during calibration are closest to the physiological reactions measured in the presence of the product.

[0011] The database thus created, in which subjective and physiological reactions are linked to emotions, serves as the basis for grouping emotions into different categories. However, this method has certain limitations. In particular, close groupings of emotions that are opposite in terms of pleasure may be observed. Furthermore, this method did not intend to use direct physiological measurements, such as electroencephalograms (EEGs), which would allow for the direct measurement of the pleasure induced by a product. It turns out that this limitation compromises the accuracy of the classification of product perception into different categories, especially for emotions such as energy and fear.Physiological reactions, such as BVP amplitude, heart rate (HR), and SCL, appear similar in response to certain opposing emotions due to the lack of a measurement for pleasure, and therefore make it difficult to distinguish between these emotions.

[0012] It also appears that the experimental protocol implemented in the method proposed by the Applicant constitutes another limitation. Indeed, in this protocol, participants had to close their eyes while an experimenter presented a bottle to their nose. This design limited the application of the process to olfactory perception alone, which made it unsuitable for products requiring tasting.

[0013] There is therefore a need to adapt the process developed by the Applicant, described above, to overcome these limitations, and in particular to adapt it to the testing of products such as food and beverages. Summary

[0014] Embodiments relate to a method for determining a theme elicited by a sensory stimulation to be characterized, comprising steps of: (a) subjecting a human subject to visual stimulation; (b) subjecting the subject to a sensory stimulation to be characterized; (c) acquiring physiological measurement signals from the participants during the submission of the visual stimulation and during the submission of the sensory stimulation to be characterized; (d) presenting the subject with an image showing a set of emojis during or after the submission of the sensory stimulation to be characterized; (e) acquiring gaze-tracking signals from the subject during the presentation of the image; (f) processing the gaze-tracking signals to determine a first emoji number fixed for the longest time by the subject, the emoji fixed for the longest time determining a group of themes likely to be elicited in the subject by the sensory stimulation to be characterized;and (g) process the physiological measurement signals acquired during the submission of visual stimulation, and during the submission of sensory stimulation to be characterized, in order to identify a theme elicited in the subject, within the determined group of themes.

[0015] By implementing various physiological measurements and analyzing eye movements while viewing an image displaying emojis, it is possible to characterize a product thematically by identifying a theme that corresponds to an emotional state elicited by sensory stimulation triggered by the product. This characterization is based not on subjective responses from the subject, but on physiological measurements that directly measure a sensation of pleasure and emotional intensity. These physiological measurements also allow for the evaluation of such a theme triggered by sensory stimulation, such as tasting food, which can simultaneously affect several senses, including taste, smell, sight, and hearing. The use of emojis eliminates language barriers and cultural specificities that can vary from one individual to another.Furthermore, the identified themes can be of all kinds, such as a type of emotion or a film genre, for example.

[0016] According to one embodiment, the processing of physiological measurements includes the use of a table listing themes and in which each theme is associated with an emoji number, and with a point defined by a pleasure score and an emotional intensity score.

[0017] The use of such a table of themes makes it easy to determine themes from physiological measurements.

[0018] According to one embodiment, the processing of the physiological measurements includes steps consisting of: determining a pleasure score and an emotional intensity score from the physiological measurements; identifying in the table all the themes corresponding to the first emoji number determined; and calculating a distance between a point defined by the determined pleasure and emotional intensity scores and each of the points associated with the themes identified in the table, the theme identified for the subject corresponding to the smallest calculated distance.

[0019] Determining scores of pleasure and emotional intensity from physiological measurements makes it possible to establish a link with the table of themes, and to easily determine a theme that allows characterizing the sensory stimulation that can act on different senses.

[0020] According to one embodiment, the eye-tracking signal processing provides a second emoji number fixed for the longest time by the participant, when another emoji has been fixed by the subject for at least 50% of the fixation time of the emoji corresponding to the first emoji number fixed for the longest time.

[0021] The possibility of providing a second emoji allows for a more nuanced characterization of the theme felt by a human subject, aroused by sensory stimulation.

[0022] According to one embodiment, the acquisition and processing of physiological measurement signals includes steps consisting of: acquiring electroencephalogram signals from electrodes placed on the subject's head, the electrodes comprising right and left pairs of frontal electrodes, during the submission of the sensory stimulation to be characterized; processing the electroencephalogram signals to determine percentage differences between the right and left electrodes of each pair of frontal electrodes; for each percentage difference, determining a pleasure index based on the membership of the percentage difference in a range of percentage values ​​from a set of percentage ranges specified for the pleasure indices; and determining a pleasure score from the pleasure indices.

[0023] The use of electroencephalogram signals advantageously allows for the evaluation of a sensation of pleasure. Furthermore, the use of percentage ranges associated with pleasure index values ​​makes it easy to evaluate this sensation of pleasure in a manner consistent with other physiological measurement methods.

[0024] According to one embodiment, the pleasure score is determined by adding the pleasure indices to a pleasure score determined according to the membership of a pleasure value indicated by the subject, in a range of pleasure values ​​of a set of pleasure ranges determined for the pleasure values.

[0025] Combining several measures with a value provided by the subject allows a subjective aspect to be taken into account, which makes it possible to determine the pleasure score more precisely.

[0026] According to one embodiment, the acquisition and processing of physiological measurement signals includes steps consisting of: acquiring electrodermal activity signals from electrodermal activity sensors placed on the subject, during the submission of visual stimulation and during the submission of sensory stimulation to be characterized; processing the electrodermal activity signals to determine mean values ​​and maximum deviations of electrodermal activity amplitude and heart rate values ​​of the subject, during a period of submission of visual stimulation and a period of submission of sensory stimulation to be characterized;determine a percentage difference in mean electrodermal activity values, a percentage difference in maximum electrodermal activity amplitude differences, and a percentage difference in heart rate values ​​between the visual stimulation submission period and the sensory stimulation submission period to be characterized; and for each percentage difference in electrodermal activity and heart rate, determine an emotional intensity index based on the percentage difference's membership in a range of percentage values ​​from a set of percentage ranges determined for the percentage difference; and determine an emotional intensity score from the emotional intensity indices.

[0027] The combined measurement of electrodermal and cardiac activity, and the comparison of measurements before and during the sensory stimulation to be characterized, makes it possible to objectively determine the emotional intensity felt by the subject. Here again, the use of a range of values ​​makes it easy to determine index values ​​and an emotional intensity score.

[0028] According to one embodiment, the emotional intensity score is determined by adding the emotional intensity indices to an emotional intensity score determined according to the membership of an emotional intensity value indicated by the subject, in a range of emotional intensity values ​​of a set of emotional intensity ranges determined for the emotional intensity values.

[0029] Combining several measures with a value provided by the subject allows a subjective aspect to be taken into account, which makes it possible to determine the emotional intensity score more precisely.

[0030] According to one embodiment, the method comprises steps consisting of: performing steps (a) to (g) for each participant in a group of participants to determine a theme for the participant; and determining a thematic characterization result of sensory stimulation based on themes determined for all participants in the group of participants.

[0031] Thus, by applying the process to a panel of participants, it is possible to characterize sensory stimulation on a thematic level, for example caused by the perception of a product.

[0032] According to one embodiment, the sensory characterization result includes a list of themes in which each theme is associated with a theme score determined according to the number of times the theme was identified for a participant in the group of participants.

[0033] The determination of a list of themes makes it possible to characterize in detail the sensory stimulation, for example caused by the perception of a product.

[0034] Embodiments may also relate to a system comprising: a computer, an eye-tracking device connected to the computer, physiological parameter measurement devices connected to the computer, and a display screen connected to the computer, the processor being configured to implement the method defined above.

[0035] According to one embodiment, the physiological parameter measurement devices include: a set of electroencephalogram electrodes, and a device for measuring electrodermal and cardiac activity. Brief description of the figures

[0036] The present invention will be better understood with the aid of the following description of exemplary embodiments with reference to the accompanying figures, in which identical reference signs correspond to structurally and / or functionally identical or similar elements.

[0037] [Fig-1] Figure 1 schematically represents an acquisition and processing of physiological measurements and responses of human subjects, according to one embodiment,

[0038] [Fig. 2] Figure 2 represents an image presented to a human subject for the determination of emotions felt by the latter, according to one embodiment,

[0039] [Fig. 3] Figure 3 is a schematic top view of an array of EEG sensors used in the method according to one embodiment,

[0040] [Fig.4] Figure 4 is a schematic front view of a device for measuring HR, SCL and BVP, used in the process according to one embodiment,

[0041] [Fig. 5] Figure 5 schematically represents functions of the acquisition and processing system, according to one embodiment,

[0042] [Fig.6] Figures 6A, 6B are representations illustrating a characterization of an emotional state caused by sensory stimulation, determined by the acquisition and processing system, according to one embodiment. Detailed description

[0043] Figure 1 represents an APS system for acquiring and processing physiological measurements and responses from human subjects in a group of human subjects participating in a product test, in order to characterize product perception. The APS acquisition and processing system comprises a PRC computer connected to a DSP display screen, an EEG EGS sensor array, a PHS device for measuring physiological parameters such as electrodermal activity and cardiac activity, and an ETR eye-tracking system.

[0044] The ETR eye-tracking system is configured to monitor and analyze a person's eye movements. It includes an optical camera that projects near-infrared light onto the cornea, thus facilitating the detection of eye position. It provides measurements indicating at any given moment what the person is focusing their gaze on.

[0045] According to one embodiment, a first type of measurement is performed by placing a test participant in front of the DSP display screen and displaying an image such as the DEM image shown in Figure 2 on this screen for a duration, for example, of 10 seconds. During the time that the participant can observe the DEM image displayed on the DSP screen, the ETR eye-tracking system is active and acquires the path of a central point observed by the participant. The DEM image displays El-El 5 emojis representing primary emotions selected based on their relevance to the type of product being tested. In the example in Figure 2, the DEM image displays approximately fifteen emojis, including: - an El emoji illustrating a neutral emotional state that can correspond to states ranging from boredom to a certain pleasant state, - an E2 emoji illustrating emotional states of surprise ranging from a state of great pleasant surprise to a state of great unpleasant surprise, passing through a neutral state of distraction, - an E3 emoji illustrating emotional states of wonder ranging from ecstasy to astonishment, - an E4 emoji illustrating emotional states of escapism ranging from seeking adventure to a state of flight, - an E5 emoji illustrating emotional states of passion ranging from elation to fervor, - an E6 emoji illustrating emotional states of feeling energy, ranging from revitalizing to stimulating energy, - an E7 emoji illustrating emotional states of curiosity ranging from enthusiasm to doubt, including interest, - an E8 emoji illustrating emotional states of disgust ranging from repugnance to boredom to ardor, - an E9 emoji illustrating unpleasant emotional states ranging from dissatisfaction to rejection, - an E10 emoji illustrating emotional states of nostalgia ranging from melancholy to reminiscence, - an Eli emoji illustrating emotional states of guilty pleasure ranging from the feeling of committing a delicious sin to mild indulgence, - an E12 emoji illustrating emotional states of sensuality ranging from eroticism to delicacy, - an El3 emoji illustrating a state of emotional relaxation ranging from well-being to calm, - an E14 emoji illustrating emotional states of tenderness ranging from intense pleasure to a simple feeling of comfort, and - an E5 emoji illustrating emotional states of joy ranging from ecstasy to serenity.

[0046] Figure 3 shows the electroencephalogram (EEG) sensor assembly in the form of an EGS headset with twelve electrodes, each capable of capturing the electrical activity of approximately one hundred neurons. The electrodes are placed on the headset to obtain information about the emotional state of the user wearing the headset on their head. These electrodes include six left frontal electrodes (F3, FP1, AF7) and six right frontal electrodes (F4, FP2, AF8), two left parietal electrodes (P3) and two right parietal electrodes (P4), two left occipital parietal electrodes (PO7) and two right occipital electrodes (PO8), and two left occipital electrodes (O1) and two right occipital electrodes (O2) (and one ground electrode (GND) in a central frontal position). Research has demonstrated the effectiveness of EEG signals in assessing an emotional index. More specifically, these studies have focused on asymmetric activity in the frontal brain (refs.

[21] ,

[22] ,

[23] ).This asymmetry involves observing differences in activity between the right and left frontal areas in the alpha band (8–12 Hz). Such differences can indicate responses such as approach or withdrawal motivations. According to the FAA frontal asymmetry theory of EEG, increased activity in the right hemisphere signifies a withdrawal motivation, while increased activity in the left hemisphere suggests an approach motivation (refs.

[24] ,

[25] ). Furthermore, attention can be measured using the theta / alpha index algorithm. This... measurement focuses on the parietal electrodes P3 and P4 for theta and the frontal electrodes F3 and F4 for alpha (ref.

[25] ).

[0047] Figure 4 shows the PHS device for measuring electrodermal activity (EDA) and cardiac activity. The PHS device can be a photoplethysmograph that uses an optical sensor to measure infrared light to detect volumetric changes in blood flow at the fingertips. In the example shown in Figure 4, the PHS device is a case that is attached by means of a strap to the index and middle fingers of the user's UHA hand. The PHS device includes, for example, an infrared light sensor (SI) that is placed on the index finger to measure electrodermal activity and an infrared light sensor (S2) that is placed on the middle finger to measure cardiac activity. The measurements taken by the PHS device can be performed, for example, with a sampling rate of approximately 64 samples per second.

[0048] Sensory information received through perceiving an odor or tasting food is processed by the olfactory system. Signals are then sent to the brain to regulate the activity of the autonomic nervous system (ANS) (ref.

[26] ). Heart rate is a well-established measure of ANS activity (refs.

[27] ,

[28] ,

[29] ,

[30] ). The heart is regulated by the ANS, particularly by its sympathetic (excitatory) and parasympathetic (inhibitory) branches. These branches exert regulatory control over heart rate by influencing the activity of the sinoatrial node, the heart's natural pacemaker. Heart rate measurement is one of the components used to assess the user's emotional arousal through sympathetic nervous system activity (ref.

[31] ).

[0049] Electrodermal activity (EDA) refers to the conductance or electrical potential of the skin, particularly the hands, which can be measured as a reflex response to emotional stimuli. EDA serves as an electrophysiological indicator reflecting the activity of the sympathetic nervous system (SNS). This activity originates from the sweat glands, which are responsible for the secretion of aqueous sweat and are distributed over the entire body surface, including the palms of the hands. While the primary function of these glands is thermoregulation, resulting in cutaneous vasodilation in response to elevated temperatures, they are also innervated by two types of neurons. Neurons using acetylcholine as a neurotransmitter are responsible for the electrodermal activity of the sweat glands in response to SNS activation.In contrast, neurons using adrenergic neurotransmitters induce vasodilation in response to temperature increases. Therefore, the phenomenon of sweating without vasodilation, called "emotional sweating," occurs due to AED in response to SNS activation (ref.

[32] ).

[0050] According to an embodiment illustrated in Figure 5, the APS system implements a method for characterizing an emotional state induced by sensory stimulation based on physiological measurements acquired during a test session involving several participants, conducted according to a precise protocol. According to one example of such a protocol, each test participant is positioned in front of a system such as the APS system, places the EGS headset on their head and the PHS device on one of their hands. Bottles containing samples of products to taste are placed in front of each participant. The ETR eye-tracking system is calibrated on the DSP display screen to track eye movements while maintaining a static head position. The signals provided by the electrodes of the EGS headset and by the PHS device are also checked and calibrated, if necessary, before starting the test.Thus, the test begins with a calibration of the ETR system, asking participants to fix their gaze for 30 seconds on a cross displayed on the DSP screen. This initial 30-second step is followed by a 5-second instruction asking participants to take a sample bottle of the product and taste its contents. Participants are then asked to fix their gaze for 10 seconds. The DEM image is then displayed for 10 seconds so that participants can visualize the emojis that best correspond to the sensations experienced while tasting the sample. This step is followed by another fixation step during which participants are again asked to fix their gaze for 10 seconds. In a final step, participants may be asked to complete a questionnaire. The questionnaire may include the following questions: 1. Which emoji do you select? 2. What pleasure value do you attribute to the product, compared to your expectations (from 0 indicating intense displeasure to 10 indicating intense pleasure)? 3. What emotional intensity value do you assign to the emotion felt when you tasted the product (from 0 indicating no emotional intensity to 10 indicating very strong emotional intensity).

[0051] Based on the pleasure and emotional intensity values ​​provided by each participant, an RP function of the PRC calculator determines pleasure (NP) and emotional intensity (El) (NEI) scores as follows:

[0052] - if the value is less than 3, the rating is set to -2, - If the value is between 3 (inclusive) and 5 (exclusive), the grade is set at -1. - If the value is equal to 5, the grade is set at 0. - if the value is between 5 (not included) and 7 (inclusive), the grade is set at 1, - if the value is greater than 7, the grade is set at 2.

[0053] The PRC computer determines several parameters based on data determined from measurements. These parameters include two parameters The main two indices are: a PS pleasure index and an EIS EI index. The PS pleasure index takes into account the evaluation of pleasure and the frontal asymmetry of alpha waves (FAA) within a 10-second window after tasting the product. The EIS EI index is determined based on the change in heart rate between the 30-second window after tasting the product and the 10-second window before tasting, the average changes, and possibly the amplitude of the EDA level before and after tasting.

[0054] To this end, throughout the test, the FP1S, FP2S, AF7S, AF8S, F3S, and F4S signals from the FP1, FP2, AF7, AF8, F3, and F4 electrodes of the EGS headset, the SS1 and SS2 signals from the SI and S2 sensors of the PHS device, and the ETC signals from the ETR eye-tracking system are recorded and processed. The FP1S, FP2S, AF7S, AF8S, F3S, and F4S signals from the EGS headset are preprocessed by an EPPS function of the PRC processor. The EPPS function samples the FP1S, FP2S, AF7S, AF8S, F3S, and F4S signals at a frequency of 256 Hz, then filters the resulting samples with a 1 to 25 Hz bandpass filter. This initial filtering is followed by a three-stage filtering process. In the first step, large-amplitude artifacts are removed using an Artifact Subspace Reconstruction (ASR) technique.In the second step, Independent Component Analysis (ICA) is applied to the filtered signals to extract the sources of the components influencing the signals. In the third step, the Multiple Artifact Rejection Algorithm (MARA) is run to automatically classify the ICA components as artifacts. This classification of independent components is performed using a pre-trained model. Components that are not related to brain activity, such as blinks or pulse rate, can thus be removed. Once the signals are filtered, various frequency calculations, such as Individualized Alpha Frequency Analysis (IAF), are applied to obtain signals related to alpha, beta, gamma, delta, and theta waves (ref.

[33] ).The signals relating to the different waves are then processed to calculate EEG indices such as frontal alpha asymmetry (FAA) following methodologies described in existing literature, and using frequency bands delimited using the Welch method to obtain a power spectral density. DP%, DA%, and DF% functions convert the measurements obtained into percentage values ​​of the FPD, AFD, and FD difference between the corresponding right and left electrodes.

[0055] Thus, the DP% function calculates a percentage FPD difference between the right and left signals from the FP1-FP2 electrode pair, the DA% function calculates a percentage AFD difference between the right and left signals from the AF7-AF8 electrode pair, and the DF% function calculates a percentage FPD difference between The right and left signals from the F3-F4 electrode pair. The FPD and AFD deviation percentages are processed to obtain FPI and AFI indices as follows:

[0056] - if the percentage difference is less than -15%, the corresponding index is set to -2, - if the percentage difference is between -15% (inclusive) and -5% (exclusive), the corresponding index is set at -1, - if the percentage difference is between -5% (inclusive) and 4% (exclusive), the corresponding index is set at 0, - if the percentage difference is between 4% and 15% (inclusive), the corresponding index is set at 1, - if the percentage difference is greater than 15%, the corresponding index is set at 2.

[0057] The percentage deviation FD is processed to obtain an index FI, as follows:

[0058] - if the percentage difference is less than -15%, the corresponding index is set at -3, - if the percentage difference is between -15% (inclusive) and -5% (exclusive), the corresponding index is set at -2,

[0059] - if the percentage difference is between -5% (inclusive) and 4% (exclusive) (including), the corresponding index is set to 0,

[0060] - if the percentage difference is between 4% and 15% (inclusive), the index the corresponding value is set to 2.

[0061] - if the percentage difference is greater than 15%, the corresponding index is set at 3.

[0062] The FPI, AFI, FI indices are then added together with the NP pleasure rating to obtain the PS pleasure score.

[0063] Electrodermal activity (EDA) and cardiac activity are analyzed by the PRC computer using SS1 signals from the SL sensors. A preprocessing function, SCP1, samples the SS1 signal at 32 Hz. The SCP1 function smooths the skin conductance level (SCL) signals related to EDA to remove small-amplitude artifacts. An SCPR function processes the output signals from SCP1 to obtain average SMV values ​​and maximum amplitude deviations (SAV) over time segments of interest, namely during the display of the cross on the DSP screen, and during or after tasting the product to be characterized. From the SMV and SAV values, the DM% and DA% functions respectively determine the percentage deviations (SCM and SCA) between the rest period and during or after tasting the product sample, average SMV values, and maximum amplitude deviations (SAV).

[0064] For heart rate analysis, an SCP2 function of the PRC calculator can also apply SS2 signal sampling processing at 32 Hz and sample smoothing to remove low-amplitude artifacts. A function HRPR processes smoothed samples to determine the time between each heartbeat, allowing the calculation of an instantaneous heart rate (HRV) during rest and after product tasting. A DH% function processes the instantaneous heart rate to obtain an HRP percentage of heart rate variation between the resting period and the periods during and after tasting the product sample.

[0065] Functions ICM, ICS process the percentage changes SCM, SCA to determine average value indices SMI and maximum amplitude deviation indices S AI, in the following manner:

[0066] - if the percentage change in the CMI is less than 0%, the average value index SMI is set at -3. - if the percentage change in the ICM is between 0% and 4% (inclusive), the average value index (SMI) is set to 0. - if the percentage change in the ICM is greater than 4%, the average value index (SMI) is set at 2. - if the percentage variation ICS is less than 8%, the maximum amplitude deviation index S AI is set to 0, - if the percentage variation ICS is between 8% and 15% (inclusive), the S AI index is set at 1, - if the percentage variation ICS is greater than 15%, the index S AI is set to 2,

[0067] An IHR function processes the percentage HRP variation of heart rate to determine an HRI index of heart rate variation, as follows:

[0068] - if the percentage variation in HRP is less than -6%, the HRI index is set at -2, - if the percentage variation in HRP is between -6% and 6% (inclusive), the HRI index is set to 0, - if the percentage of variation HRP is greater than 6%, the HRI index is set at 2.

[0069] The SMI, SAI mean and maximum amplitude deviation indices and the HRI index are added together with the EI NEI score to obtain the EI EIS score.

[0070] Thus, in the calculations of the PS pleasure and EIS EI scores, greater importance is given to physiological measures than to the VP, VEI values ​​of pleasure and EI provided by the participant.

[0071] The ETR system records the ETC coordinates of the gaze of the right and left eyes on the screen at a sampling frequency of 60 Hz. An ETPS function of the PRC computer receives the ETC coordinates and determines, at each instant, a moving average of the X coordinates of the right and left eyes, as well as the Y coordinates of the right and left eyes, to obtain the average coordinates (Xavg and Yavg) of the central point of gaze. An ETPR function determines the duration The duration of eye contact with each emoji is calculated based on the trajectory of the participant's gaze, with emojis arbitrarily numbered from 1 to 15. For this purpose, the IZ of each emoji (E1-E15) is considered to encompass the entire display area of ​​the emoji in the DEM image and extend a few millimeters beyond the emoji. The fixation times for each emoji are cumulative. In one embodiment, the IZ of an emoji is delimited by a circle centered on the emoji and having a diameter equal to the distance to either of two adjacent emojis.

[0072] The ETPR function can also determine the number of emoji fixations per emoji, provided that an emoji fixation is counted when the time the gaze is held in the emoji's area of ​​interest (IZ) exceeds a certain minimum threshold value. This minimum threshold value is, for example, set at 200 ms. A total number of emoji fixations can thus also be determined. The ETPR function also determines a hesitation score (HI) based on the total time spent in the emoji's areas of interest (IZ) divided by the total number of emoji fixations.

[0073] An EMP function processes the fixation durations lEe of each emoji e to determine whether one or more emojis are relevant to the participant's emotional experience. Since each emoji is numbered, the EMP function determines a first EPI number for the most relevant emoji, i.e., the emoji that was fixed for the longest fixation time. A second most relevant emoji number, EP2, can also be determined. The second most relevant emoji is the one that was fixed by the participant for at least 50% of the time spent in the area of ​​interest of the most relevant emoji. If the hesitation score (HI) is less than or equal to 1, the second relevant emoji is taken into account.

[0074] The Pleasure Score (PS), the Emotional Intensity Score (EIS), and the first most relevant EPI emoji number, as well as optionally the second most relevant EP2 emoji number, are provided to an SCPR function of the PRC calculator along with an ET1 table. The ET1 table comprises a list of emotions, each emotion being associated in the table with an emoji number and Pleasure Score and EI values. The emotions in the ET1 table are also grouped by emotional universe. The ET1 table can be defined according to the following example:

[0075] [Tables 1] Universe Emotion Emoji Number Pleasure Score Balanced Neutral Neutral 1 0 0 Satisfaction Emotion Joy 2 2 0 Serenity 2 4 -3 Ecstasy 2 8 8 Curiosity 10 4 4 Interest 10 1 0 Enthusiasm 10 8 8 Emotional Tranquility Tenderness 3 4 0 Comfort 3 0 -5 Intense Pleasure 3 8 8 Calm 4 2 0 Relaxation 4 4 -3 Well-being 4 8 4 Sensual Temptation Sensuality 5 4 4 Delicacy 5 0 -3 Eroticism 5 8 8 Indulgent Temptation Guilty Pleasure 6 4 0 Mild Indulgence 6 0 -5 Delicious Sin 6 8 6 Emotional Excitement Pleasant Surprise 15 3 4 Pleasant Astonishment 15 8 8 Wonder 14 4 2 Astonishment 14 0 -4 Ecstasy 14 8 8 Escape 13 4 0 Flight 13 2 -3 Adventure 13 8 8 Passion 12 4 2 Ardor 12 6 4 Exaltation 12 8 8 Stimulating Energizer 11 4 2 Stimulating 11 2 0 Revitalizing 11 8 8 Nostalgic Reflection Nostalgia 7 2 0 Reminiscence 7 5 6 Melancholy 7 -3 0 Negative Aversion Disgust 9 -4 0 Boredom 9 0 -6 Repugnance 9 -8 8 Dissatisfaction 8 0 -6 Displeasure 8 -4 0 Rejection 8 -8 8 Unpleasant Surprise 15 -2 4 Unpleasant Astonishment 15 -6 8 Doubtful 10 -3 0 Fleeting Attention Distraction 15 0 0

[0076] In this example, the ET1 table associates each emotion with an emoji number from 1 to 15, and pleasure and EI indices ranging from -8, corresponding to a feeling of pleasure or very negative emotional intensity, to +8, corresponding to a feeling of pleasure or very positive emotional intensity, with 0 corresponding to a neutral feeling. The ET1 table thus defines a set of points in a three-dimensional space (x=emoji number, y=pleasure score, z=EI score).

[0077] Based on the EPI emoji number, and possibly EP2, and the PS pleasure and EIS emotional intensity scores of each participant and the ET1 table, the SCPR function performs a classification for each participant. The first classification places the participant's reaction in one of the emotion universes specified in the ET1 table, while the second classification places the participant's emotion within that universe.

[0078] To this end, the SCPR function considers the plane (x=EPl, y, z) of this three-dimensional space, corresponding to the participant's EPI emoji number, and determines the distance between the participant's point (x=EPl, y=PS, z=EIS) and each of the points defined in table ET1, located in this plane. The emotion corresponding to the point defined by table ET1, located closest to the participant's point p in the plane corresponding to the participant's EPI emoji number, determines the EMlp emotion felt by the latter. The SCPR function then identifies in table ET1 an emotional universe Ulp for the participant p, based on the EMlp emotion.

[0079] If a second EP2 emoji number was selected by the ETC signal analysis, a second EM2p emotion is determined for participant p in the same way, by considering the point in table ET1 closest to the participant's point in the plane (x=EP2, y, z) defined by the second EP2 emoji number. This emotion classification method thus makes it possible to identify a precise emotion that best corresponds to what the participant felt. The SCPR function then identifies an emotional universe U2p for participant p in table ET1, based on the EM2p emotion.

[0080] According to one embodiment, if a participant point is too far from the other points in the same emotional group corresponding to the emoji number EPI or EP2, this means that the emotional group in question does not correspond to the participant's reaction. For this purpose, a maximum distance threshold value can be applied. This maximum distance threshold value can be defined for each emotional group. For example, it can correspond to the maximum distance existing between the points in the emotional group. Thus, the distance thresholds can be defined as specified in the following table:

[0081] [Tables2] Emoji Number Maximum Distance Threshold 0 0 1 4.00 2 11.70 3 15.26 4 8.06 5 13.60 6 13.30 7 10.00 8 16.12 9 16.12 10 13.60 11 10.00 12 7.21 13 12.53 14 14.42 15 14.00

[0082] The analysis is then redirected to another emotion group (or emoji number). This other group may be the one corresponding to the second EP2 emoji selected by the ETC signal analysis. If there is no second emoji, then no emotion is assigned.

[0083] Next, the PRC processor determines a score for each emotion listed in table ET1, based on the number of times the emotion (EM1p, EM2p) is identified among the participants in the participant group. To this end, the PRC processor can apply a weighting of 2 when only one EM1p emotion is identified in participant p, and weightings that sum to 2 when two EM1 and EM2 emotions are identified. These weightings can be set to 1 and 1, or 1.25 and 0.75, or determined based on the respective fixation times of the two corresponding EM1 and EM2 emojis.

[0084] The PRC processor also determines a score for each emotional universe listed in Table ET1, based on the number of times the emotional universe (Ulp, U2p) is identified among the participants in the participant group. For this purpose, emotion weighting rules can be applied to the emotional universes. The score for each emotional universe can also be determined by summing the scores obtained for the emotions belonging to the emotional universe as specified in Table ET1.

[0085] The emotion and emotional universe scores obtained for a product can be presented as a map of emotions and emotional universes. In this representation, the emotional universes are represented as rectangles whose area corresponds to the score associated with the emotional universe. Then, each emotion in each emotional universe is also represented as a rectangle whose area corresponds to the score associated with the emotion, this rectangle being inscribed within the rectangle corresponding to the emotional universe.

[0086] Figures 6A and 6B show examples of map representations of score lists obtained for a product. The representation in Figure 6A includes a large rectangle EEX corresponding to the universe of emotional excitement, a smaller rectangle ETQ corresponding to the universe of emotional tranquility, an even smaller rectangle IBT corresponding to the universe of stimulating energy, four rectangles NAV, IDT, ECT, and NRF of the same size, each even smaller, corresponding to the universes of negative aversion, indulgent temptation, emotional satisfaction, and nostalgic reflection, and finally an even smaller rectangle FAT corresponding to the universe of fleeting attention. The rectangle EEX includes rectangles of different sizes. The surface area includes a rectangle WD corresponding to the emotion of wonder, a rectangle ES corresponding to the emotion of escapism, a rectangle EY corresponding to the emotion of ecstasy, and two rectangles AS and PS of the same size corresponding to the emotions of astonishment and passion. The rectangle ETQ comprises rectangles of different sizes, including a rectangle TD corresponding to the emotion of tenderness, and two rectangles CF and IP of the same size corresponding to the emotions of comfort and intense pleasure. The rectangle IBT comprises a rectangle EG corresponding to the emotion of energizing. The rectangle NAV comprises a rectangle DS corresponding to the emotion of dissatisfaction. The rectangle IDT comprises a rectangle GP corresponding to the emotion of guilty pleasure. The rectangle ECT comprises a rectangle JY corresponding to the emotion of joy. The rectangle FAT comprises a rectangle DT corresponding to the emotion of distraction.

[0087] The representation in Figure 6B comprises a large rectangle ETQ corresponding to the universe of emotional tranquility, two smaller rectangles EEX, FCT corresponding to the universes of emotional excitement and emotional satisfaction, an even smaller rectangle FAT corresponding to the universe of fleeting attention, two even smaller rectangles NAV, IDT of the same size corresponding to the universes of negative aversion and indulgent temptation, and finally an even smaller rectangle RTM corresponding to the universe of sensual temptation. The rectangle ETQ comprises rectangles of different areas, including a rectangle TD corresponding to the emotion of tenderness, a rectangle CM corresponding to the emotion of calm, and two rectangles RX, WB of the same size corresponding to the emotions of relaxation and well-being.The EEX rectangle comprises rectangles of varying sizes, including a GW rectangle representing the emotion of detachment, an ES rectangle representing the emotion of escapism, and an EY rectangle representing the emotion of ecstasy. The ECT rectangle comprises a JY rectangle representing the emotion of joy. The FAT rectangle comprises a DT rectangle representing the emotion of distraction. The IBT rectangle comprises an EG rectangle representing the emotion of energizing. The NAV rectangle comprises a DS rectangle representing the emotion of dissatisfaction. The IDT rectangle comprises a GP rectangle representing the emotion of guilty pleasure. The RTM rectangle comprises a DL rectangle representing the emotion of delicacy.

[0088] It will be evident to those skilled in the art that the present invention is susceptible to various embodiments and applications. In particular, the invention is not limited to determining a list of emotions felt by participants in a product test. On the contrary, the present invention can be applied more generally to determining themes, which can be of all kinds, such as, for example, a film genre (visual and auditory), or a musical genre. (auditory), a product feel (tactile), a video game concept (visual, auditory), or even a vehicle interior concept.

[0089] Furthermore, it is not necessary to use a theme table such as the table 1. Indeed, other data structures can easily be developed to determine a theme from physiological measurements. For example, a list of film genres (action, adventure, comedy, thriller, horror, romance, etc.) can be associated with physiological measurements. Furthermore, it is also possible to use other types of physiological measurements, such as respiratory rate, restlessness, facial expression, skin temperature, etc. Similarly, other types of scores can also be determined from physiological measurements, bearing in mind that these scores are highly dependent on the type of physiological measurement.

[0090] Methods for identifying a theme other than that described above by calculating distances between points can also be implemented.

[0091] Methods for determining scores other than the classification methods using ranges of values ​​each associated with a score value, described above, can also be implemented. In addition, it may be envisaged to implement classifications not on percentages of variation, but directly on the values ​​of the physiological measurements, and by fixing the range of possible values ​​of the index resulting from this classification.

[0092] Furthermore, taking into account designated emojis or values ​​of pleasure and emotional intensity, provided by the participants, is not necessary, given that these indications simply allow for refining the result.

[0093] The method as described above can be applied to a neurofeedback system, for example, to guide a person in decision-making, or to determine a person's emotional state, for example, in terms of anxiety or stress. It is therefore not essential to apply the method to a panel of participants and to provide a general result based on the individual results obtained for the participants. References

[0094] 1. Bradley, Margaret M., and Peter J. Lang. 1994. “Measuring Emotion: The Self- Assessment Manikin and the Semantic Differential”. Journal of Behavior Therapy and Experimental Psychiatry 25(1):49-59. doi: 10.1016 / 0005-7916(94)90063-9.

[0095] 2. Bensafi, M., C. Rouby, V. Farget, B. Bertrand, M. Vigoureux, and A. Holley. 2002. « Psychophysiological Correlates of Affects in Human Olfaction ». Neurophysiologie Clinique / Clinical Neurophysiology 32(5):326-32. doi: 10.1016 / 80987-7053(02)00339-8.

[0096] 3. Jack, Rachael E., Oliver G. B. Garrod, Hui Yu, Roberto Caldara, et Philippe G. Schyns. 2012. « Facial Expressions of Emotion Are Not Culturally Universal ». Proceedings of the National Academy of Sciences 109(19):7241-44. doi: 10.1073 / pnas. 1200155109.

[0097] 4. Warrenburg, S. 2005. « Effects of Fragrance on Emotions: Moods and Physiology ». Chemical Senses 30(Supplement l):i248-49. doi: 10.1093 / chemse / bjh208.

[0098] 5. Porcherot, Christelle, Sylvain Delplanque, Sophie Raviot-Derrien, Bénédicte Le Calvé, Christelle Chrea, Nadine Gaudreau, et Isabelle Cayeux. 2010. « How Do You Feel When You Smell This? Optimization of a Verbal Measurement of Odor-Elicited Emotions ». Food Quality and Preference 21(8):938-47. doi: 10.1016 / j.foodqua 1.2010.03.012.

[0099] .6. Churchill, Anne, et John Behan. 2010. « Comparison of Methods Used to Study Consumer Emotions Associated with Fragrance ». Food Quality and Preference 21(8):1108-13. doi: 10.1016 / j.foodqual.2010.07.006.

[0100] 7. Vidal, Leticia, Gaston Ares, et Sara R. Jaeger. 2016. « Use of Emoticon and Emoji in Tweets for Food-Related Emotional Expression ». Food Quality and Preference 49:119-28. doi: 10.1016 / j.foodqual.2015.12.002.

[0101] 8. Jaeger, Sara R., Armand V. Cardello, et Howard G. Schutz. 2013. « Emotion Questionnaires: A Consumer-Centric Perspective ». Food Quality and Preference 30(2):229-41. doi: 10.1016 / j.foodqual.2013.05.015.

[0102] 9. Thomson, David M. H., et Christopher Crocker. 2015. « Application of Conceptual Profiling in Brand, Packaging and Product Development ». Food Quality and Preference 40:343-53. doi: 10.1016 / j.foodqual.2014.04.013.

[0103] 10. Truss. 2004. « Truss 2004 An educational companion to eats shoots & leaves.pdf ».

[0104] 11. Cowie, R., E. Douglas-Cowie, N. Tsapatsoulis, G. Votsis, S. Kollias, W. Fellenz, et J. G. Taylor. 2001. « Emotion Récognition in Human-Computer Interaction ». IEEE Signal Processing Magazine 18(1):32-80. doi: 10.1109 / 79.911197.

[0105] 12. Huang, Albert H., David C. Yen, et Xiaoni Zhang. 2008. « Exploring the Potential Effects of Emoticons ». Information & Management 45(7):466-73. doi: 10.1016 / j.im.2008.07.001.

[0106] 13. Huffaker, David A., et Sandra L. Calvert. 2006. « Gender, Identity, and Language Use in Teenage Blogs ». Journal of Computer-Mediated Communication 10(2):00-00. doi: 10.111 l / j,1083-6101.2005.tb00238.x.

[0107] 14. Wolf, Alecia. 2000. « Emotional Expression Online: Gender Différences in Emoticon Use ». CyberPsychology & Behavior 3(5):827-33. doi: 10.1089 / 10949310050191809.

[0108] 15. Barros, Catarina, Ana Rita Perdra, Adriana Sampaio, Ana Bujân, et Diego Pinal. 2022. « Frontal Alpha Asymmetry and Négative Mood: A Cross-Sectional Study in Older and Younger Adults ». Symmetry 14(8):1579. doi: 10.3390 / syml4081579.

[0109] 16. Palacios-Ibânez, Almudena, Javier Marin-Morales, Manuel Contero, et Mariano Alcaniz. 2023. « Predicting Decision-Making in Virtual Environments: An Eye Movement Analysis with Household Products ». Applied Sciences 13(12):7124. doi: 10.3390 / appl3127124.

[0110] 17. Rojas, Juan-Carlos, Javier Marin-Morales, José Manuel Ausin Azofra, et Manuel Contero. 2020. « Recognizing Decision-Making Using Eye Movement: A Case Study With Children ». Frontiers in Psychology 11:570470. doi: 10.3389 / fpsyg.2020.570470.

[0111] 18. Wedel, Michel, Rik Pieters, et Ralf Van Der Lans. 2023. « Modeling Eye Movements During Decision Making: A Review ». Psychometrika 88(2):697-729. doi: 10.1007 / sl 1336-022-09876-4.

[0112] 19. Lim, Jia Zheng, James Mountstephens, et Jason Teo. 2020. « Emotion Récognition Using Eye-Tracking: Taxonomy, Review and Current Challenges ». Sensors 20(8):2384. doi: 10.3390 / s20082384.

[0113] 20. Zhu, Jianping, Lizhen Ji, et Chengyu Liu. 2019. « Heart Rate Variability Monitoring for Emotion and Disorders of Emotion ». Physiological Measurement 40(6):064004. doi: 10.1088 / 1361-6579 / abl887.

[0114] 21. Golnar-Nik, Pamaz, Sajjad Farashi, et Mir-Shahram Safari. 2019. « The Application of EEG Power for the Prédiction and Interprétation of Consumer Decision-Making: A Neuromarketing Study ». Physiology & Behavior 207:90-98. doi: 10.1016 / j.physbeh.2019.04.025.

[0115] 22. Hyun, Yong J., Byungho Park, Sanghak Lee, Sungjun Park, et Moon-Yong Kim. 2018. « A Neuro-Scientific Analysis of Affect and Cognition in Consumer Response to the Marketing Stimulus - EEG Analysis of the Relationship between Prefrontal Hemispheric Lateralization and Occipital Alpha Blocking in Ad Exposure ». SSRN Electronic Journal, doi: 10.2139 / ssrn.3211155.

[0116] 23. Lin, Meng-Hsien (Jenny), Samantha N. N. Cross, William J. Jones, et Terry L. Childers. 2018. « Applying EEG in Consumer Neuroscience ». European Journal of Marketing 52(1 / 2):66-91. doi: 10.1108 / EJM-12-2016-0805.

[0117] 24. Huffmeijer, Renske, Lenneke R. A. Alink, Mattie Tops, Marian J. Bakermans- Kranenburg, et Marinus H. Van Uzendoorn. 2012. « Asymmetric Frontal Brain Activity and Parental Rejection Predict Altruistic Behavior: Modération of Oxytocin Effects ». Cognitive, Affective, & Behavioral Neuroscience 12(2):382-92. doi: 10.3758 / s 13415-011-0082-6.

[0118] 25. Sandoval, Pamela Simon, et Jésus Garcla-Madariaga. 2024. « Impact of Emotional Appeal on Non-profit Advertising: A Neurophysiological Analysis ». Journal of Consumer Behaviour 23(1):203-17. doi: 10.1002 / cb.2168.

[0119] 26. Masuo, Yoshinori, Tadaaki Satou, Hiroaki Takemoto, et Kazuo Koike. 2021. « Smell and Stress Response in the Brain: Review of the Connection between Chemistry and Neuropharmacology ». Molécules 26(9):2571. doi: 10.3390 / molecules26092571.

[0120] 27. Bensafi, M. 2002. « Autonomie Nervous System Responses to Odours: The Rôle of Pleasantness and Arousal ». Chemical Senses 27(8):703-9. doi: 10.1093 / chemse / 27.8.703.

[0121] 28. Kadohisa, Mikiko. 2013. « Effects of Odor on Emotion, with Implications ». Frontiers in Systems Neuroscience 7. doi: 10.3389 / fnsys.2013.00066.

[0122] 29. Marci, Cari D., Debra M. Glick, Rebecca Loh, et Darin D. Dougherty. 2007. « Autonomie and Prefrontal Cortex Responses to Autobiographical Recall of Emotions ». Cognitive, Affective, & Behavioral Neuroscience 7(3):243-50. doi: 10.3758 / CABN.7.3.243.

[0123] 30. Rodriguez, S., M. Fernandez, A. Cepedabenito, et J. Vila. 2005. « Subjective and Physiological Reactivity to Chocolaté Images in High and Low Chocolaté Cravers ». Biological Psychology 70(1):9-18. doi: 10.1016 / j.biopsycho.2004.10.001.

[0124] 31. Ghamari, Mohammad. 2018. « A Review on Wearable Photoplethysmography Sensors and Their Potential Future Applications in Health Care ». International Journal of Biosensors & BioeleXctronics 4(4). doi: 10.15406 / ijbsbe.2018.04.00125.

[0125] 32. Grapperon, J. 2012. « La mesure de la réaction électrodermale ».

[0126] 33. Moya, Ingrit, Jésus Garcla-Madariaga, et Marîa-Francisca Blasco. 2020. « What Can Neuromarketing Tell Us about Food Packaging? » Foods 9(12):1856. doi: 10.3390 / foods9121856.

[0127] 34. Garcla-Madariaga, Jésus, Maria-Francisca Blasco Lôpez, Ingrit Moya Burgos, et Nuria Recuero Virto. 2019. « Do Isolated Packaging Variables Influence Consumers’ Attention and Preferences? » Physiology & Behavior 200:96-103. doi: 10.1016 / j.physbeh.2018.04.030.

[0128] 35. Plutchik, Robert. 1982. « A Psychoevolutionary Theory of Emotions ». Social Science Information 21(4-5):529-53. doi: 10.1177 / 053901882021004003.

[0129] 36. Zheng, Wei-Long, Jia-Yi Zhu, et Bao-Liang Lu. 2019. « Identifying Stable Patterns over Time for Emotion Récognition from EEG ». IEEE Transactions on Affective Computing 10(3):417-29. doi: 10.1109 / TAFFC.2017.2712143.

Claims

Demands

1. A method for determining a theme elicited by a sensory stimulus to be characterized, comprising the steps of: (a) subjecting a human subject to visual stimulation; (b) subjecting the subject to a sensory stimulus to be characterized; (c) acquiring physiological measurement signals (FP1S, FP2S, AF7S, AF8S, F3S, F4S, SS1, SS2) from the participants during the submission of the visual stimulation and during the submission of the sensory stimulus to be characterized; (d) presenting the subject with an image (DEM) showing a set of emojis (E1-E15) during or after the submission of the sensory stimulus to be characterized; (e) acquiring eye-tracking signals (ETC) from the subject during the presentation of the image;(f) process eye-tracking signals to determine the first emoji number (EPI) fixed for the longest time by the subject, the longest-fixed emoji determining a group of themes likely to be aroused in the subject by the sensory stimulation to be characterized; and (g) process physiological measurement signals acquired during the submission of the visual stimulation, and during the submission of the sensory stimulation to be characterized, to identify a theme aroused in the subject, within the determined group of themes.

2. A method according to claim 1, wherein the processing of physiological measurements (FP1S, FP2S, AF7S, AF8S, F3S, F4S, SS1, SS2) comprises the use of a table (ET1) listing themes and in which each theme is associated with an emoji number, and with a point defined by a pleasure score and an emotional intensity score.

3. A method according to claim 2, wherein the processing of the physiological measurements (FP1S, FP2S, AF7S, AF8S, F3S, F4S, SS1, SS2) comprises the steps of: determining a pleasure score (PS) and an emotional intensity score (EIS) from the physiological measurements; identifying in the table (ET1) all themes corresponding to the first emoji number (EPI) determined; and calculate a distance between a point defined by the determined pleasure (PS) and emotional intensity (EIS) scores and each of the points associated with the themes identified in the table, the theme identified for the subject corresponding to the smallest calculated distance.

4. 4. A method according to any one of claims 1 to 3, wherein eye-tracking signal processing (ETC) provides a second emoji number (EP2) fixed longest by the participant, when another emoji (E1-E15) has been fixed by the subject for at least 50% of the fixation time of the emoji corresponding to the first emoji number (EPI) fixed longest.

5. 5. A method according to any one of claims 1 to 4, wherein the acquisition and processing of physiological measurement signals (FP1S, FP2S, AF7S, AF8S, F3S, F4S, SS1, SS2) comprises steps of: acquiring electroencephalogram signals (FP1S, FP2S, AF7S, AF8S, F3S, F4S) from electrodes (FP1, FP2, AF7, AF8, F3, F4) placed on the subject's head, the electrodes comprising right and left pairs of frontal electrodes (FP1-FP2, AF7-AF8, F3-F4), during the submission of the sensory stimulation to be characterized; processing the electroencephalogram signals to determine percentage differences (FPD, AFD, FD) between the right and left electrodes of each of the pairs of frontal electrodes; for each percentage deviation, determine a pleasure index (FPI, AFI, FI) based on the membership of the percentage deviation in a range of percentage values ​​from a set of percentage ranges specified for the pleasure indices;and determine a pleasure score (PS) from the pleasure indices.

6. 6. A method according to claim 5, wherein the pleasure score (PS) is determined by adding the pleasure indices (FPI, AFI, FI) to a pleasure score (NP) determined according to the membership of a pleasure value (VP) indicated by the subject, in a range of pleasure values ​​of a set of pleasure ranges determined for the pleasure values.

7. 7. A method according to any one of claims 1 to 6, wherein the acquisition and processing of physiological measurement signals comprises steps consisting of: acquiring electrodermal activity signals (SS1, SS2) from electrodermal activity sensors (SI, S2) placed on the subject, during the submission of visual stimulation and during the submission of sensory stimulation to be characterized; process electrodermal activity signals to determine mean values ​​(SMV) and maximum amplitude deviations (SAV) of electrodermal activity and heart rate values ​​(HRV) of the subject, during a period of visual stimulation submission and a period of sensory stimulation submission to be characterized; determine a percentage deviation (SCM) of mean electrodermal activity values, a percentage deviation (SCA) of maximum amplitude deviations of electrodermal activity and a percentage deviation (HRP) of heart rate values, between the period of visual stimulation submission and the period of sensory stimulation submission to be characterized;and for each percentage difference in electrodermal activity and heart rate, determine an emotional intensity index (EMI, S AI, HRI) based on the percentage difference belonging to a range of percentage values ​​from a set of percentage ranges determined for the percentage difference; and determine an emotional intensity score (EIS) from the emotional intensity indices.

8. 8. A method according to claim 7, wherein the emotional intensity score (EIS) is determined by adding the emotional intensity indices (SMI, S AI, HRI) to an emotional intensity score (NEI) determined based on the membership of an emotional intensity value (VEI) indicated by the subject, in a range of emotional intensity values ​​from a set of emotional intensity ranges determined for the emotional intensity values.

9. 9. A method according to any one of claims 1 to 8, comprising steps of: performing steps (a) to (g) for each participant in a group of participants to determine a theme for the participant; and determining a thematic characterization result of sensory stimulation as a function of the themes determined for all participants in the group of participants.

10. 10. A method according to claim 9, wherein the result of sensory characterization comprises a list of themes in each theme is associated with a theme score determined based on the number of times the theme was identified for a participant in the group of participants.

11. 11. System comprising: a computer (PRC), an eye tracking device (ETR) connected to the computer, physiological parameter measurement devices (EGS, PHS) connected to the computer, and a display screen (DSP) connected to the computer, the processor being configured to implement the method according to any one of claims 1 to 10.

12. 12. System according to claim 11, wherein the physiological parameter measurement devices comprise: an electroencephalogram (EEG) set of electrodes, and a device (PHS) for measuring electrodermal and cardiac activity.

Images